among infants and young children than among other age groups; therefore, this pervasive change in the regime of mortality had particular impact on parents in the process of forming their families. Many parents now in their 30's and 4CTs are supporting more children than their parents did at a similar age. This change is not due to noticeably higher fertility or earlier marriage; it is simply a reflection of the fact that more children from the postwar generation have survived. At the social level the increase in child survival and in the speed at which families grow results in a more youthful population; there are fewer adults to work for each child they must feed, clothe, shelter, train, and equip for adult life. This rise in the child dependency rate implies that parents are today less able than were their parents to spread the consumption demands of their offspring over their productive lifetimes.
When childbearing is concentrated in a woman's first 10 to 15 years of marriage, and the overlapping period of child dependency is prolonged by increasing child investments, these large concurrent demands on parents' resources depress any residual earnings left for physical savings (26). Indeed, in developed countries, such as the United States, household physical savings rates fluctuate widely from periods of substantial dissaving during family formation and child schooling to a period of substantial parental savings after children achieve economic independence (14, 16, 17, 22, 29). With the emergence of a similar concentrated childbearing pattern in some low income countries, it is to be expected that differences in household physical savings rates between parents of different ages (stages of family formation) will grow more marked. There are as yet, however, few reliable data on household savings for low income countries by parent age and family size and composition. Thus it is difficult to estimate the effects changing patterns of family formation and age structure have had on savings.*
death rates. In any case, it is widely believed today that modern medicine and sanitation are not capable of inducing further large-scale reductions in death rates. The preconditions are not yet present for controlling the endemic nonmicrobial, diarrhea-pneumonia diseases that still account for a major share of the deaths in low income countries. To cut death rates further, there will have to be a widespread improvement in nutrition particularly among young children, and perhaps also a substantial advance in general living levels. Sec (24, 27, 28).
'•"Fluctuations in physical savings by age of parents might differ between developed and less developed countries for at least two reasons. First, the availability of consumer debt funding (and our frequent current expense treatment of consumer durables) permits young parents to acquire consumption loans that are not likely to be available in a low income country. Therefore, the dissavings rates noted among young parents in the developed countries might not occur. On the other hand, the concentrated childbearing pattern has gradually evolved in countries like the United States, whereas the pattern has emerged more rapidly in the low income countries. If the pattern were anticipated by parents, they might save more before starting to have children, or space their births more uniformly to spread out resource demands. However, in low income countries parents could hardly anticipate the drop in child death rates and the rapid extension of education. They may be subjected to relatively more severe resource binds and thus be compelled to reduce consumption levels or borrow in some fashion against the future when